In this week's #TidyTuesday video, I go over an interesting package called Modeltime that applies the Tidy framework to #Forecasting. Using a bike dataset, I show how to use differencing to make a stationary dataset. Using #Tidymodels, I show why it is important to split the data by its time component instead of the traditional random sampling. I then go over the main functions of modeltime and create an ARIMA, Prophet, and TSLM model. I show how to quickly validate forecast results and the important metrics in forecasting. I then refit the models to make future forecasts and show how models are sensitive to its input data.
Connect with me on LinkedIn: / andrew-couch
Q&A Submission Form: forms.gle/6EzU4GCR9VnJx8gg7
Code for this video: github.com/andrew-couch/Tidy-...
TidyTuesday: github.com/rfordatascience/ti...
PC Setup (Amazon Affiliates)
Keyboard: amzn.to/3Bbbk3T
Mouse: amzn.to/3BcRGVo
Microphone: amzn.to/3ePo9JS
Audio Interface: amzn.to/3qTAmjz
Webcam: amzn.to/3L9Ql6j
CPU: amzn.to/3qGa6Zu
GPU: amzn.to/3DnhMHL
RAM: amzn.to/3LdTxh7
Негізгі бет Ғылым және технология TidyTuesday: Tidy Forecasting with Modeltime
Пікірлер: 25